Comparison of MUSIC algorithms using gradient‐ and integral‐type mode vectors for direction of arrival and angular spread estimation

Abstract

SUMMARYA generalized mode vector that simultaneously uses the differential coefficient component of the mode vector has been proposed for estimating the direction of arrival of incident waves having angular spread. In this paper, the authors first examined this generalized mode vector and proposed a gradient-type MUSIC algorithm that uses the differential coefficient component of the mode vector independently for estimating the direction of arrival. Then they used a computer simulation in which the waves having angular spread were considered to be a group of scattered waves to compare this algorithm with regard to the performance for estimating the direction of arrival with an integraltype MUSIC algorithm that uses simple integration, which they proposed separately. To improve the estimation precision degradation that occurs when the signal correlation is low among scattered waves that exist within the angular spread when using the integral-type MUSIC algorithm, the authors proposed a combined-type MUSIC algorithm that incorporated the gradient-type MUSIC algorithm in the integral-type MUSIC algorithm and used this to examine the performance for estimating the direction of arrival and angular spread again. The results of the comparison of the gradient-type MUSIC algorithm and integral-type MUSIC algorithm confirmed that the two algorithms have conflicting properties related to the signal correlation among scattered waves that exist within the angular spread. Also, the investigation related to the combined-type MUSIC algorithm indicated that estimation precision degradation that occurs when the signal correlation is low among scattered waves when using the integral-type MUSIC algorithm was reduced as expected.